First lady Michelle Obama and her two daughters were televised wearing J. Crew's designs as President Barack Obama was sworn into office in 2009. A day later, the store's website crashed.
Of course, companies can prevent a website crash by increasing capacity when they anticipate a flood of traffic -- but predicting an unexpected IT event such as the one J. Crew experienced isn't always possible.
For an enterprise with multi-tiered applications with load-balancing on the front-end and a database server and Web servers on the back-end, adding an additional server often requires several changes.
IT must configure the Web server, change the database server's security credentials and add the Web server to the load balancer -- a timely process.
A tool like Chef automates this process so the application scales up seamlessly. With automation, infrastructure becomes testable, versioned and repeatable. Tasks are codified, so an enterprise could add 20 extra Web servers if the application suddenly took off, like J. Crew's Website did.
How to handle an unexpected IT event
Chris McLeandirector of design and professional engineer, Markley Group
The unexpected event might have a far greater impact on the business than on IT, so IT must prioritize the business's needs, according to Clive Longbottom, co-founder and service director at Quocirca, a U.K.-based research and analysis company.
"The first rule is to … plan for as much as possible," Longbottom said. "Create business continuity and recovery plans that match with the business' priorities -- and then test them."
Businesses must define a contingency plan or a business continuity plan that relies as little as possible on technology, falling back on manual processes where necessary. In the case of an unexpected IT event, this enables some level of business to continue while the issue is fixed.
Companies should also consider a contingency plan. This way, it can respond systematically to an unexpected event.
However, contingency plans are uncommon, especially for businesses with colocation providers, because the business might not understand what IT needs.
A good colocation provider will help plan and make recommendations to businesses on how to efficiently use their data center, said Chris McLean, director of design and professional engineer at the Markley Group, a wholesale colocation, hosting and cloud provider in Boston.
The divide between IT pros and the business side of the house is a major concern in the data center that can impact how to prevent or handle an unexpected event.
"Translating an IT person's needs is hard with everyone else in the data center," McLean said.
Business continuity can only be maintained when IT and business are on the same page; when all levels of the data center are blended, you won't have any surprises.
"There is a conflict between software development and the DevOps community," said Chris Crosby, CEO of Dallas-based Compass Datacenters. "It's important to bridge the gap between all areas of expertise."
Regardless of how well integrated your IT staff is, your data center is susceptible to failure. The entire team should have a uniform method of operation in order to safely and effectively manage an unexpected event.
Make sure you have a strong team behind you that you can trust and rely on.
"You have to know your limits and have a good team," McLean said. "When you select a colocation provider, they are an extension of your staff. Go to a place that offers what you need and you can rely on in emergencies."